F98 GB model
The study was approved by the Ghent University ethics committee for animal experiments (ECD 12/28-A2). All animals were kept and handled according to the European guidelines (2010/63/EU) and housed under environmentally controlled conditions (12 h normal light/dark cycles, 20 °C – 24 °C and 40–70% relative humidity) with food and water ad libitum.
The GB F98 rat model was developed as described by Bolcaen et al. [26]. In summary, F98 rat GB cells were cultured as monolayers for three weeks. 54 Female Fisher F344 rats (Charles River®) were anesthetized with ketamine/xylazine (i.p., 4/3; 0.13 mL/100 g), and immobilized using a stereotactic frame (Model 902 Dual Small Animal Stereotaxic frame, Kopf®). After shaving, the head was swabbed with betadine and the skull was exposed through a longitudinal 1 cm scalp incision. Using a diamond drill (Dremel®), a 1 mm hole was made through the skull (2.5 mm lateral to the bregma in the right frontal hemisphere). A stereotactically guided 1 mL insulin needle was inserted at a depth of 3 mm and 5 μL of cell suspension containing 20,000 F98 cells in phosphate-buffered saline was delivered. This cell suspension was injected using a microsyringe pumpcontroller (Micro 4TM, World Precision Instruments, Sarasota, USA) over a 2-min period. The syringe was slowly withdrawn 1 min post inoculation and the incision was closed with bone wax (Aesculap AG®) and sutured.
MRI
Eight days post inoculation small animal MRI was performed on a 7 T system (PharmaScan 70/16, Bruker, Ettlingen, Germany) to confirm tumor growth. Rats were anesthetized with 2% isoflurane mixed with oxygen administered at a flow rate of 0.3 L/min. To enable the injection of a gadolinium containing contrast agent (Dotarem®, Guerbet, 2 mmol/kg), a 30-Gauge needle connected to a 60 cm long PE tube was placed intravenously in the lateral tail vein. A rat brain surface coil (Rapid Biomedical, Rimpar®, Germany) was applied around the head of the animal followed by positioning of the bed in a 72 mm whole body transmitter coil (Rapid Biomedical, Rimpar®, Germany). A T2-weighted scan (T2w SE RARE, TR/TE 3661/37.1 ms, 109 μm isotropic in plane resolution, 4 averages, TA 9′45″) was performed to localize the tumor. If tumor was visually confirmed, a 5 min T1w scan was obtained after injection of the gadolinium containing contrast agent.
PET imaging
PET images were acquired on a small animal PET scanner (FLEX Triumph II, TriFoil Imaging®, Northridge, CA, USA). A 30 min PET scan was obtained with [18F]FET or [18F]FAZA (±37 MBq; 30 min or 2 h post-injection, respectively). This time frame allowed to acquire the MRI scans during tracer uptake. Reconstruction of the PET scans was done by a 2D Maximum Likelihood Expectation Maximization (MLEM) algorithm (LabPET Version 1.12.1, TriFoil Imaging, Northridge, CA, USA) applying 50 iterations and a using voxel size of 0.5 × 0.5 × 1.157 mm3 into 200 × 200 × 64 matrix.
Treatment groups
Different therapy groups were defined based on the definition of the target volumes for brain tumor irradiation. For the MRI based RT group (n = 5), the isocenter was placed in the middle of the contrast enhancement on T1-weighted (T1w) MRI (Fig. 2). The [18F]FET-PET based RT (n = 6) and [18F]FAZA-PET based RT (n = 4) groups received an additional radiation boost of 5 Gy delivered to the region either with maximum [18F]FET or [18F]FAZA PET tracer uptake, respectively.
All therapy groups received concomitant treatment consisting of intraperitoneal (IP) injections of TMZ on five consecutive days, starting on the same day as the RT [27, 28]. Therefore, 29 mg/kg TMZ (Sigma-Aldrich®) was dissolved in saline with 25% dimethylsulfoxide.
Finally, a control group (n = 5) received no RT and only control IP injections with an equal amount of dimethylsulfoxide and saline on five consecutive days.
Animals with MRI-confirmed GB-tumor were randomly assigned to the abovementioned groups.
Radiation treatment
Radiotherapy treatment was performed using a small animal radiation research platform (SARRP, Xstrahl®, Surrey, UK). MRI-based RT in the F98 GB rat model was already optimized by our research group [23, 29]. A gadolinium-based contrast agent was intravenously injected, followed by fixation of the animal on a multimodality bed. Two water filled capillaries were used as markers and were placed on the right side and in the middle of the rat head. After acquiring contrast-enhanced T1w spin-echo scans, the rat was transported to the four-axis robotic positioning table of the SARRP, while fixed on the multimodality bed. A treatment planning CT was obtained using the following acquisition parameters: 50 kV tube voltage, 600 μA tube current, 360 projection acquired over 360 degrees using a 1 mm Al-filter, resulting in a total acquisition time of 1 min. The acquired projection data were reconstructed using an isotropic voxel size of 0.2 mm. After importing the planning CT into the treatment planning software (Muriplan, Version 2.0.6, Xstrahl®, UK), manual segmentation was performed to distinguish air, soft tissue and bone. Co-registration with the MRI was done manually using the capillary markers and the skull. Based on the contrast-enhanced T1w MRI scan the isocenter of the radiation bundle was set in the middle of the tumor region. A single dose of 20 Gy was delivered by applying three non-coplanar arcs (120°) using a 5 × 5 mm2 collimator to include minor position changes of the rat head during execution of the treatment.
For the PET/MRI-based RT planning, co-registration of PET and planning CT was performed using PMOD (version 3.31, PMOD technologies, Ltd). First, the contrast-enhanced T1w MRI scan was co-registered with the planning CT as described above. Then, the PET image was co-registered with the planning CT. Finally, the coordinates of the isocenters based on 1) contrast enhancement on the T1w MRI scan and 2) based on maximum tracer uptake in the PET image were determined in PMOD and transferred to Muriplan. The isocenter indicated on the MRI scan received a single dose of 20 Gy using a 5 × 5 mm2 collimator as described above, while the isocenter based on maximum PET tracer uptake received an additional 5 Gy boost using a 1 × 1 mm2 collimator. The methodology has been described in detail in a recently published Jove Movie [23].
Chemotherapy
Concomitant chemotherapy was given on five consecutive days starting on the same day as the RT. Therefore, IP injections of 29 mg/kg TMZ (Sigma-Aldrich®) dissolved in saline with 25% dimethylsulfoxide were performed. The control group received IP injections with an equal amount of dimethylsulfoxide and saline on five consecutive days [27, 28].
Histological characterization
At the end of the experiment or when the humane endpoints were reached (> 20% weight loss, tumor volume > 40% of total brain volume based on MRI or signs of ataxia) rats were euthanized by an IV injection of pentobarbital (120 mg/kg). From five animals the brain was isolated, dissected, immersed in 4% paraformaldehyde for 24 h and embedded in paraffin. Then, the brain was partly sectioned into 5 μm slices and stained with hematoxylin and eosin (HE) in order to histologically characterize the tumor.
Comparison of [18F]FET and [18F]FAZA PET
The uptake of both tracers was studied in the same F98 GB rat (n = 3). Additionally, the pre-RT PET scans of the [18F]FET-PET based RT (n = 6) and [18F]FAZA-PET (n = 4) based RT groups were analyzed. Using PMOD, rigid body transformations were applied to co-register the PET and MRI scans. Volumes of interest (VOI) were drawn manually and included the contrast-enhanced region on the T1w MRI scans. Cubic VOIs of 2 × 2 × 2 mm3 located in the contralateral region were used as a reference (background region). Tracer uptake in the VOI at each time frame was converted to a standardized uptake value (SUV) according to the following formula:
SUV = (Radioactivity in VOI expressed in Bq/ml / injected activity in Bq) x body weight in g.
Injected activity was corrected for radioactive decay and residual activity in the syringe. Standardized uptake values (SUVmean and SUVmax) and tumor-to-background ratios (TBRmean and TBRmax) were calculated.
Dose volume histograms
To compare the dose of MRI-based RT with PET/MRI-based sub-volume boosting, dose volume histograms (DVH) were analyzed. Using Muriplan, the DVHs were determined within the overlapping volume of the three rotating 5 × 5 mm2 bundles using Muriplan. The average, maximum and minimum dose, as well as the D2-, D50-, and D90-values, were determined and compared between different RT treatment plans. Dx stands for the dose that x % of the tissue volume received.
Assessment of therapy response
Tumor growth was evaluated by obtaining sequential MRI scans 2, 5, 9, 12, 14, and 16 days after initiating the treatment. Tumor volumes were measured by manually outlining the tumor on individual slices of contrast-enhanced T1w images using PMOD.
Statistical analysis
Statistical analyses were done using SPSS software (SPSS Statistics 23 software). If applicable, a Mann-Whitney U test was applied. A probability value of p < 0.05 was considered statistically significant. Values are presented as mean ± SD.